crash on TS-264

[McSim85]

Hey hey,

Thanks for developing such a nice exporter.
It seems like the exporter crashes when I open the page http://192.168.88.23:9094/metrics

./qnapexporter
2023/10/19 09:28:52 Listening to HTTP requests at :9094
2023/10/19 09:29:00 Reading environment...
2023/10/19 09:29:00 Hostname: MEDIA, err=<nil>
2023/10/19 09:29:00 Retrieving QTS version
2023/10/19 09:29:00 Retrieved getsysinfo path: "/sbin/getsysinfo"
2023/10/19 09:29:00 Retrieved sysdhnum: 4
2023/10/19 09:29:00 Retrieved sysfannum: 1
2023/10/19 09:29:00 Retrieved volCount: 4
2023/10/19 09:29:00 Retrieved vol_desc "[Volume SSD, Pool 1]", parsed to "SSD"
2023/10/19 09:29:00 Retrieved volume "SSD" vol_fs "EXT4"
2023/10/19 09:29:00 Retrieved volume "SSD" vol_totalsize "532.82 GB"
2023/10/19 09:29:00 Retrieved volume "SSD" vol_status "Ready"
2023/10/19 09:29:00 Retrieved vol_desc "[Volume backups, Pool 2]", parsed to "backups"
2023/10/19 09:29:01 Retrieved volume "backups" vol_fs "EXT4"
2023/10/19 09:29:01 Retrieved volume "backups" vol_totalsize "990.60 GB"
2023/10/19 09:29:01 Retrieved volume "backups" vol_status "Ready"
2023/10/19 09:29:01 Retrieved vol_desc "[Volume Volume, Pool 2]", parsed to "Volume"
2023/10/19 09:29:01 Retrieved volume "Volume" vol_fs "EXT4"
2023/10/19 09:29:01 Retrieved volume "Volume" vol_totalsize "1.01 TB"
2023/10/19 09:29:01 Retrieved volume "Volume" vol_status "Ready"
2023/10/19 09:29:01 Retrieved vol_desc "[Volume Family, Pool 2]", parsed to "Family"
2023/10/19 09:29:01 Retrieved volume "Family" vol_fs "EXT4"
2023/10/19 09:29:01 Retrieved volume "Family" vol_totalsize "3.48 TB"
2023/10/19 09:29:02 Retrieved volume "Family" vol_status "Ready"
2023/10/19 09:29:02 Found volumes [{0 EXT4 SSD Ready 0 5.7211111866368e+11} {1 EXT4 backups Ready 0 1.0636486508544e+12} {2 EXT4 Volume Ready 0 1.11050674405376e+12} {3 EXT4 Family Ready 0 3.82630046466048e+12}]
2023/10/19 09:29:02 Retrieved sysvolinfo
2023/10/19 09:29:02 Retrieved hal_app path: "/sbin/hal_app"
2023/10/19 09:29:02 Retrieving QM2 enclosures
2023/10/19 09:29:02 Retrieving network interfaces in "/sys/class/net"...
2023/10/19 09:29:02 Retrieving devices in "/dev"...
2023/10/19 09:29:02 Found devices: [nvme0n1 nvme1n1 sda sdb]
2023/10/19 09:29:02 Retrieving dm-cache devices...
2023/10/19 09:29:02 Found cache clients: [vg1-tp1_tierdata_2 vg2-tp2_tierdata_2]
2023/10/19 09:29:02 Found cache volumes: [vg256-lv256	(253:3)]
2023/10/19 09:29:02 Connecting to UPS daemon
panic: runtime error: index out of range [1] with length 1

goroutine 113 [running]:
github.com/pedropombeiro/qnapexporter/lib/exporter/prometheus.(*promExporter).getDmCacheStatsMetrics(0xc000320540)
	/github/workspace/lib/exporter/prometheus/disk.go:108 +0x545
github.com/pedropombeiro/qnapexporter/lib/exporter/prometheus.fetchMetricsWorker(0xc000043690?, 0xc0000437d0?, 0xd, 0xc000154c60?)
	/github/workspace/lib/exporter/prometheus/prometheus.go:166 +0x78
created by github.com/pedropombeiro/qnapexporter/lib/exporter/prometheus.(*promExporter).WriteMetrics
	/github/workspace/lib/exporter/prometheus/prometheus.go:126 +0x1fc

QNAP model is TS-264

[SkyeYoung]

I also encountered this problem. After looking at the code, it seems that the returned data is somewhat different on different devices.

qnapexporter/lib/exporter/prometheus/disk.go

Lines 88 to 111 in 7d7a127

	func (e *promExporter) getDmCacheStatsMetrics() ([]metric, error) {
	if len(e.dmCacheClients) == 0 {
	return nil, nil
	}
	args := append([]string{"status", "--noflush"}, e.dmCacheClients...)
	lines, err := utils.ExecCommandGetLines("dmsetup", args...)
	if err != nil {
	return nil, fmt.Errorf("get dm-cache status (dmsetup %s): %w", args, err)
	}
	metrics := make([]metric, 0, len(lines)+6*2)
	for index, line := range lines {
	tokens := strings.SplitN(line, " ", 13)
	cache := e.dmCacheClients[index]
	allocationRatioStr := strings.TrimSpace(tokens[3])
	allocationTokens := strings.SplitN(allocationRatioStr, "/", 2)
	attr := fmt.Sprintf("device=%q", cache)
	metrics = appendFloatMetric(metrics, "node_flashcache_cached_blocks", allocationTokens[0], 1, "", "")
	metrics = appendFloatMetric(metrics, "node_flashcache_total_blocks", allocationTokens[1], 1, "", "")
	metrics = appendFloatMetric(metrics, "node_dmcache_used_bytes_total", allocationTokens[0], 1024*1024, attr, "Number of blocks resident in the cache")
	metrics = appendFloatMetric(metrics, "node_dmcache_bytes_total", allocationTokens[1], 1024*1024, attr, "Total number of cache blocks")
	}

Here the string with index 3 is taken directly, but it seems that on my device it should be 4

[SkyeYoung]

I've pushed a PR to try to fix this, maybe you can also try compiling one yourself and give it a try.